A population-based case-control epidemiologic study will be conducted to investigate risk factors associated with the incidence of second primary breast cancer. Incident cases of second primary breast cancer will be identified through the Connecticut Tumor Registry and review of pathology records at Connecticut hospitals. The cases, who will number approximately 360, will be compared to an equal number of controls selected randomly from women who have developed a first primary but not a second and who are alive at the time that the second primaries develop in the cases. Interviews will be conducted using a structured questionnaire administered in person by trained interviewers. The questionnaire will include items covering reproductive history, family history of cancer, use of estrogens, socio-demographic factors, and relevant medical conditions and procedures. Medical information concerning the characteristics of the first primary cancers of both cases and controls will be obtained from hospital records. This information will include stage of disease, hormone receptor levels, size of tumor, and type of treatment. An independent pathologic review will be conducted for distinguishing second primaries from metastases and for classifying the histologic features of the first primary cancers in a uniform manner. Multiple logistic regression analysis will be performed in order to assess the effects of each factor while taking all other factors into account. The results should permit improved prediction of which women are at high risk for second primary breast cancer. This information will have important implications for clinical care and may help to elucidate further the etiology of breast cancer in general.